Fast-Path Nodes for Seamless Packet Workload Migration
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing virtualization-based services face challenges in managing large-scale packet processing workloads across multiple nodes in cloud provider networks, particularly in migrating workloads without disrupting existing packet flows and ensuring seamless performance, especially when handling millions of packets per second.
Innovation Solution
A multi-layer packet processing service with isolated node groups and a system for migrating packet processing workloads between these groups, utilizing fast-path and exception-path nodes to manage packet rewriting and state information replication, ensuring continuous operation and minimizing disruptions during migration.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If workload migration is performed between isolated node groups, then system scalability and load balancing are improved, but packet flow continuity and connection stability deteriorate due to potential connection resets and packet drops
Solution Approach 1:
The system performs preliminary actions by replicating flow state information from the source node group to the destination node group before actually migrating the workload. This advance replication ensures that when migration occurs, the destination group already has the necessary state information to handle packets seamlessly, preventing connection resets and maintaining flow continuity throughout the migration process.
Solution Approach 2:
The patent introduces fast-path nodes as intermediaries that facilitate communication and state information transfer between isolated node groups during migration. These intermediary nodes receive packets from clients, determine the appropriate destination node group, and forward packets while maintaining flow state consistency, thereby enabling workload migration without disrupting existing connections.
2Reliability
If fast-path nodes are used as intermediaries for workload migration, then migration seamlessness and packet flow continuity are improved, but system complexity and infrastructure requirements worsen
Solution Approach 1:
The fast-path nodes perform multiple functions: they handle normal packet processing, manage flow state information replication, determine destination node groups for migration, and forward packets during migration. By making these nodes multi-functional, the system achieves seamless migration without adding separate dedicated migration infrastructure, thereby managing complexity while maintaining reliability.
Solution Approach 2:
The fast-path nodes autonomously determine whether to forward packets to source or destination node groups based on flow state information and migration status. This self-service capability eliminates the need for complex external control mechanisms during migration, allowing the system to maintain seamless operation while managing complexity through decentralized decision-making at the node level.
3Measurement precision
If flow state information is replicated between node groups during migration, then packet processing accuracy and flow state consistency are improved, but network bandwidth consumption and replication overhead worsen
Solution Approach 1:
The system replicates flow state information selectively rather than universally. Flow state information is replicated only for specific flows that are undergoing migration or may be affected by migration, and only to the extent necessary for maintaining consistency. This localized approach ensures accuracy where needed while minimizing unnecessary bandwidth consumption for flows that don't require replication.
Solution Approach 2:
The patent implements partial replication of flow state information, replicating only the essential state data needed to maintain packet processing consistency during migration, rather than replicating complete flow states or all possible information. This partial action approach achieves sufficient consistency to prevent packet drops while reducing the bandwidth overhead compared to full replication.
Data Source
AI summary
A control plane server of a packet processing service assigns a first node group comprising exception-path nodes of the service to a network interface of a first application. Nodes of the assigned node group provide packet rewriting rules used by fast-path nodes of the service to direct requests of the application. In response to detecting that a workload migration criterion has been met, the control plane server initiates a migration workflow of the interface. During the workflow, flow state information of a packet flow is replicated at nodes of the first node group, sent from the first node group to a second node group via a fast-path node, and replicated at nodes of the second node group. Connections used for the application requests remain operational. After the migration workflow completes, nodes of the second node group provide packet rewriting rules for directing requests of the application.